Know any good legibility tests for typefaces?

[Double Elephant]

This topic was imported from the Typophile platform

The tests need to assess the legibility (& perhaps readability) of typefaces for print-based instructional texts (not continuous).

[Dunwich Type]

Assessing legibility is pretty simple, just look it over and make sure that similar glpyhs have distinct forms. As for readability, just print yourself a sample and start reading.

[Nick Shinn]

No matter how legible a typeface is, it's quite easy for a poor typographer to make it unreadable.

[crossgrove]

Is there something about the instructions that requires speed reading? Conventional typefaces like Times, Minion, Garamond, Charter, Georgia, Dante, Clifford, Hoefler Text, Constantia, Galliard, Electra, Goudy Oldstyle, etc. are all adequate for instructions, tests, forms, etc. as long as they aren't mangled (negative tracking, inadequate linespacing, light inking, artificially condensed). There are better ones and of course much worse ones (Helvetica, Bodoni, display typefaces), but is there some critical reason the typefaces have to have some legibility rating? What kind of texts are you preparing?

I ask because there is no such standard test; it's a huge bugaboo in the type and typography world, and unfortunately it's all decided ad hoc. your best bet right now is actually to make a "safe" or conventional choice and not mess with the type too much; set it well. If you want assistance in choosing highly legible typefaces, be prepared either for a cacophonous disagreement, or else very subjective choices from whoever you ask.

You may have to design the test yourself, if this information is required. Contact me offline for suggestions.

[ebensorkin]

Carl, I for one would love to have a list your personal top 3, 5, or 10 from you if you are willing.

[crossgrove]

Here, or shall I e-mail you?

[ebensorkin]

Whichever you prefer. Thanks very much!

[russellm]

I know of some bad legibility tests.

I think that type designers typographers and graphic designers should get busy and work on some way of gauging relative legibility & readability before optometrists do it.

I already have to follow guidelines that are set by people who think they know about typography because they can misuse the word Kerning with complete confidence.

-=®=-

[Nick Shinn]

I think that type designers typographers and graphic designers should get busy and work on some way of gauging relative legibility & readability before optometrists do it.

You aren't really suggesting that we voluntarily assimilate with the Borg?

[SuperUltraFabulous]

Karmina

[russellm]

You aren’t really suggesting that we voluntarily assimilate with the Borg?

No. but, graphic designers may find their hands tied, in signage if not elsewhere by accessibility standards that are set by the typographically ignorant.

What's a Borg? :p

-=®=-

[ebensorkin]

"Borg" is a Star Trek reference

http://en.wikipedia.org/wiki/Borg_(Star_Trek)

Nick are you really against developing something to help test this properly? Or do you think it can't be done properly?

[Kevin Larson]

Optometrists and reading psychologists have been measuring aspects of reading performance for quite a while. This recent paper by a team of optometrists and reading psychologists uses a variety of methodologies: reading speed, distance threshold, preference, and eye strain reports. In this paper they aren’t testing fonts, only different versions of ClearType rendering:

Sheedy, J., Y.-C. Tai, et al. (2008). "ClearType sub-pixel text rendering: Preference, legibility and reading performance." Displays 29(2): 138-151.

Mary Dyson tells me that she is planning on using this paper in a typeface design seminar at the University of Reading because it is a nice introduction to different testing methods.

While optometrists and reading psychologists come from a different academic background than typographers, I don’t think it’s useful to describe them as the Borg. There is a common goal of providing a good reading experience and everyone stands to gain from cooperation.

Cheers, Kevin

p.s. Since I work for Microsoft I am expecting Borg jokes. I’m ready.

[Nick Shinn]

Nick are you really against developing something to help test this properly? Or do you think it can’t be done properly?

I'm against the productization of readability and the hyping of such type products with scientific or statistical terminology.
The effect of this is to derogate the acumen of type designers and typographers, by replacing their subtle creativity and design skills with principles that are simplistic and banal. So the client doesn't defer to the typographer's judgement, but wants hard proof. That would be nice, but typographic design, whether of typefaces or page layouts, doesn't work like that. Typography is design. Glyphs are drawn by hand.

Linotype was the Borg of the newspaper industry in the mid 20th century. The Legibility Series of types it developed played a significant role in making newspaper typography the stagnant backwater of the graphic design world, until the advent of DTP.

**

There's scientific research, which should test how people read. And there's product development, which tests how products perform. The efficacy of a design cannot be determined before it is marketed, and in many cases it creates its own demand (hence innovation and progress).

About a century ago, some clever people began applying scientific principles to advertising (Daniel Starch, for instance, and Claude Hopkins). But the furthest they have ever gotten down that road (apart from the subliminal detour) is market research. For actually coming up with creative, you want Kirk on the job, not One of Nine.

There is a common goal of providing a good reading experience and everyone stands to gain from cooperation.

Some however may resist such platitudes, prefering to be independent and not become assimilated into the optimum reading experience.

Performance principles do have a place in typography (for road signage and special needs accessibility), but typefaces should not be tagged with readability standards, which is where this is leading.

[Dan Gayle]

< nerd> not One of Nine < /nerd>

[crossgrove]

"I’m against the productization of readability and the hyping of such type products with scientific or statistical terminology.
The effect of this is to derogate the acumen of type designers and typographers, by replacing their subtle creativity and design skills with principles that are simplistic and banal. So the client doesn’t defer to the typographer’s judgement, but wants hard proof."

And yet this happens anyway; see Read Regular, Green Font, Tiresias, etc. etc. It's going to happen anyway, without scientific "proof". People still want the proof, even if it can't be had. What to do? I agree that it reduces too much, too simplistically, but it would sometimes be handy to be able to point to something easy for non-specialists to understand. There certainly is no consensus among typographers.

[ebensorkin]

Nick I think you make some great points about the culture of type and business.

Despite that I think it would be great if really useful thorough research was done. It could be useful for Scientists, useful for Users, and useful for font makers. I think all of that & more could definitely be done if there was funds and a will. But I am not holding my breath! I am happy and in a way grateful that there is position at MS for Kevin! Were that there were more!

But the reality is that doing this research isn't a big priority in our Civ. We are very rich civilization and we could be doing much much much more. But we don't. This is party because we already seem to have an embarrassment of riches when it comes to fonts - and it's snowballing! And partly because the purpose of one font vs. the other is, so much of the time that of sensibility & tone served with the usual portion of utility; rather than sheer utility or with utility coming as a first or dominant priority. And because of that I think that the Nick Shinns of the world need not fear too much. Maybe this is a mistaken impression. Maybe there is more pressure as regards utility than I know. Feel free to correct me.

The efficacy of a design cannot be determined before it is marketed

This seems mistaken to me. You can learn things ahead of release of a font about it's utility. What you can't know is how well a face is going to sell. Which thing are you getting at here?

Some however may resist such platitudes, etc...

I agree that being assimilated into "group thought" would be a bad plan. But that isn't science. The conflation of Marketing & Science isn't helpful, even if one thing may trail on shortly after the other. Once Kevin's work published is in a peer reviewed scientific journal it is cooperative. And when MS decides that Kevin is allowed publish MS funded research then; to that extent, they are genuinely cooperating and contributing to the common good. The key word being extent.

Also in fairness to Kevin you can't expect his boss' priorities to be the general welfare of western civ or civ in general much as they might like to spin it that way – and have. If it can't be used for competitive advantage then it might not be worth it for them. They need to justify what they do in service to MS. Maybe there can be a bit of overlap as regards the world but that has to be secondary. By the same token you can't expect Nick to point to a competitor's typeface on Typophile. Obviously sheer scale means that it isn't the same. But I think it's the same kind of thing.

I also think there is also always going to be a cultural/experiential overlay to the results despite even an ideal testing process*. Data collected in OH might be different from NYC might be different from Germany & so on. So it may (or may not be) that you can get solid usable results but the practical application is questionable.

I feel am not contradicting myself here. Or shading things grey. I am just attempting to parcel things out.

For actually coming up with creative...

I think "Creative" isn't for the purposes of this discussion one thing. It is at least two. It is sensibility & tone rather and utility.** Obviously you are right when it comes to selecting for sensibility & tone. But utility is a different animal. It is true that if you can know the purpose a font will be put to, LCD screens or Newsprint or whatever you can use your eyes and go with your educated impression. And that works very well. If it didn't we wouldn't see the advances we have. What I reject in the abstract is the idea that Science can't be effectively added to the team as it were. The "unfair" advantage is simply that MS can afford the services of a scientist. But compared to dropping fonts into the market for free I suspect this is a minor advantage.

hyping of such type products with scientific or statistical terminology

Here I agree. Anybody can come up with a list of the most readable fonts ( or a best font) & then drop a white paper that is supposed to prove it. Okay not anybody. But is done in many industries and constantly. You have to be skeptical. And people are, more & more. But being critical is science. What sucks is when they decide that it's a good plan to pitch science along with it's superficial trappings. Don't conflate scientific process & hype. Hype doesn't have to attend Science. And when it does they can be separated. It isn't fun, but it can be done.

Also from what I can tell people have been less & less inclined to accept hype with re: to fonts. I don't think that the reception given to ClearType has been rapturous. It wasn't as bad as the din about Vista but still...

I agree that it reduces too much, too simplistically, but it would sometimes be handy to be able to point to something easy for non-specialists to understand.

Science is complicated and grey much of the time. It is really really rare and wonderful when real science can be presented as something easy to understand. Like "you need enough vitamin C to avoid Scurvey so get this citrus in your diet". I don't have much hope that this "something" will be produced for use with clients. And I am even less ceratin that they would make good use of it. I am much more hopeful that many small somethings might be available to type designers*** to make use of. And that we could be more effective at marketing.

And yet this happens anyway

"There is one born every Minute" comes to mind. This is why Marketing exists. I agree with Carl. I think you have to shrug it off basically. His "what to do" is certainly more pithy!

* Assuming such a thing could exist which it cannot.
** Maybe you don't mean to include utility but given the context of your use of it it seemed like you did.
*** Or type making companies

For those of you that made it this far - sorry about this overly long post.

[Kevin Larson]

Some however may resist such platitudes

The vast majority of reading research is devoted to understanding how people read, and is entirely agnostic to the concerns of typography. There are several conferences and journals devoted to understanding word recognition, reading acquisition, dyslexia, and reading comprehension where type is not considered an interesting variable. The researchers at all of these conferences find it perfectly acceptable to use Courier as their main typeface. That’s what we’ve achieved by resisting cooperation.

[Jackie Frant]

Nick: No matter how legible a typeface is, it’s quite easy for a poor typographer to make it unreadable.

A truer thought has not been spoken in years.

One test on readability we use to do for text in the shop was a relatively simple one. After the first repro came out - we turned it upside down. We made sure it had a nice balance of gray, and that there were no rivers. It was truly easier to see this upside down, because "your" proofing eyes did not have to concentrate on the words - typos - and other weird set behavior. (Oh, the days before WYSIWYGs)

[glytch]

The "readability" of *texts* can be measured by the Flesch index. This is, on the face of it, a "common-sense" approach to decoding text. But any count of words per line and characters per word, while appearing to be objective analysis, can be "gamed" by a writer who is trying to increase readability.

Imagine how typographic/marketing/design decisions could be "justified" (no pun intended) by calibrating such seemingly objective features as length of descenders, character count, stroke weight contrast, etc., then tossing in a couple of constants and passing the whole thing off as objective "analysis."

Designers already invoke such visual characteristics as "greater x-height" or "serifs" as primary signifiers of legibility. But, at what point will a further increase in x-height begin to *reduce* legibility? How long can a serif be, before it *interferes* with the reading process? Perhaps the latest fashion trend in fonts (popularity) is a more reliable predictor of legibility than any structural element inherent to the characters themselves; if it is read more often, mustn't it be more legible? Legibility of printed instructions would seem to vary more in direct proportion to time available for reading, as well as the visual acuity and emotional state of the reader, not unlike many other forms of perception.

[ebensorkin]

if it is read more often, mustn’t it be more legible?

No. See Arial. Arial is with us as a result of a business decision and prior cultural history. Ubiquity is not anything like a useful measure.

Perhaps the latest fashion trend in fonts (popularity) is a more reliable predictor of legibility than any structural element inherent to the characters themselves

This is just silly because it mixes impact that voice flavor or style have on popularity with the effects that the sheer utility of the font have. I would be lovely ( in a creepy sort of way ) if the world was so simple but it isn't.

It's cool that you are thinking about it though.

[John Boardley]

@hellbox
You might like to read Gerard Unger's While You're Reading.

I do think that trying to make a science of legibility is pretty pointless. No doubt someone with a BSc will concoct a grand scientific method. However, I suggest that a better test is the reader's opinion (and your own, if you trust it). Typography is an art, not a science; and I'm much more interested in the opinions of good typographers, than even the most informed opinions of the world's most eminent scientists.

[dberlowgone]

"Typography is an art, not a science; and I’m much more interested in the opinions of good typographers, than even the most informed opinions of the world’s most eminent scientists."

Can I have your baby?

"Sheedy, J., Y.-C. Tai, et al. (2008). "

Where, please? Must keep up to data.

Cheers!

[crossgrove]

"I do think that trying to make a science of legibility is pretty pointless. No doubt someone with a BSc will concoct a grand scientific method. However, I suggest that a better test is the reader’s opinion (and your own, if you trust it). Typography is an art, not a science; and I’m much more interested in the opinions of good typographers, than even the most informed opinions of the world’s most eminent scientists."

I think the proposition is that "the world's most eminent scientists" would build their tests around.... The Reader's Opinion. Determining which existing typefaces are more legible has less to do with scientists vapidly collecting abstract metrics, and more to do with testing typefaces in use: reading speed, comprehension, comfort over long reading times, eye fatigue. It doesn't presuppose dissecting or destroying anything, and it isn't necessarily divorced from what typographers "know" about type. The trick is to get typographers involved in designing tests. That's what Kevin is on to. One of the most frustrating flaws in research to date is wretchedly bad typeface selection. Courier, Curlz, Arial Condensed, Helveeta, etc. are not appropriate for testing in extended reading.

[Nick Shinn]

Eventually, the ergonomic project bogs down in demographics.
You can't design the one-size-fits-all product, you have to target a sector.
That's a distinction you can dispense with if your company owns most of the market.